diff --git a/crates/float/src/lib.rs b/crates/float/src/lib.rs
index deaa7776..e261a8d5 100644
--- a/crates/float/src/lib.rs
+++ b/crates/float/src/lib.rs
@@ -12,7 +12,7 @@ use revm::interpreter::interpreter::EthInterpreter;
 use revm::primitives::{address, fixed_bytes};
 use revm::{Context, MainBuilder, MainContext, SystemCallEvm};
 use std::cell::RefCell;
-use std::ops::{Add, Div, Mul, Sub};
+use std::ops::{Add, Div, Mul, Neg, Sub};
 use std::thread::AccessError;
 use thiserror::Error;
 
@@ -232,6 +232,16 @@ impl Float {
         })
     }
 
+    pub fn inv(self) -> Result<Self, FloatError> {
+        let Float(a) = self;
+        let calldata = DecimalFloat::invCall { a }.abi_encode();
+
+        execute_call(Bytes::from(calldata), |output| {
+            let decoded = DecimalFloat::invCall::abi_decode_returns(output.as_ref())?;
+            Ok(Float(decoded))
+        })
+    }
+
     pub fn abs(self) -> Result<Float, FloatError> {
         let Float(a) = self;
         let calldata = DecimalFloat::absCall { a }.abi_encode();
@@ -303,6 +313,20 @@ impl Div for Float {
     }
 }
 
+impl Neg for Float {
+    type Output = Result<Self, FloatError>;
+
+    fn neg(self) -> Self::Output {
+        let Float(a) = self;
+        let calldata = DecimalFloat::minusCall { a }.abi_encode();
+
+        execute_call(Bytes::from(calldata), |output| {
+            let decoded = DecimalFloat::minusCall::abi_decode_returns(output.as_ref())?;
+            Ok(Float(decoded))
+        })
+    }
+}
+
 #[cfg(test)]
 mod tests {
     use super::*;
@@ -531,6 +555,64 @@ mod tests {
         }
     }
 
+    #[test]
+    fn test_minus_format() {
+        let float = Float::parse("-123.1234234625468391".to_string()).unwrap();
+        let negated = float.neg().unwrap();
+        let formatted = negated.format().unwrap();
+        assert_eq!(formatted, "123.1234234625468391");
+
+        let float = Float::parse(formatted).unwrap();
+        let negated = float.neg().unwrap();
+        let formatted = negated.format().unwrap();
+        assert_eq!(formatted, "-123.1234234625468391");
+
+        let float = Float::parse("0".to_string()).unwrap();
+        let negated = float.neg().unwrap();
+        let formatted = negated.format().unwrap();
+        assert_eq!(formatted, "0");
+    }
+
+    proptest! {
+        #[test]
+        fn test_minus_minus(float in arb_float()) {
+            let negated = float.neg().unwrap();
+            let renegated = negated.neg().unwrap();
+            prop_assert!(float.eq(renegated).unwrap());
+        }
+    }
+
+    proptest! {
+        #[test]
+        fn test_inv_prod(float in reasonable_float()) {
+            let zero = Float::parse("0".to_string()).unwrap();
+            prop_assume!(!float.eq(zero).unwrap());
+
+            let inv = float.inv().unwrap();
+            let product = (float * inv).unwrap();
+            let one = Float::parse("1".to_string()).unwrap();
+
+            // Allow for minor rounding errors introduced by the lossy
+            // `inv` implementation. We consider the property to
+            // hold if the product is within `±1e-37` of 1.
+
+            let eps = Float::parse("1e-37".to_string()).unwrap();
+            let one_plus_eps = (one + eps).unwrap();
+            let one_minus_eps = (one - eps).unwrap();
+
+            let within_upper = !product.gt(one_plus_eps).unwrap();
+            let within_lower = !product.lt(one_minus_eps).unwrap();
+
+            prop_assert!(
+                within_upper && within_lower,
+                "float: {}, inv: {}, product: {} (not within ±ε)",
+                float.show_unpacked().unwrap(),
+                inv.show_unpacked().unwrap(),
+                product.show_unpacked().unwrap(),
+            );
+        }
+    }
+
     proptest! {
         #[test]
         fn test_abs_no_minus_sign(float in reasonable_float()) {